--- On Mon, 1/17/11, Lynn W Deffenbaugh (Mr) wrote:
> I wasn't here then, but I suspect the > complaints were targeted at the > data source, but were in fact the problem of IGate > operators that > configured their IGates to transmit alerts for a much > larger area than > their reasonable RF footprint. It's easier to shut > down the source than > to "fix" a few (tens? hundreds?) of IGates that might not > have even > supported sufficient filtering capability to even stem the > tide of messages. > > The weather alerts as specified at http://www.aprs-is.net/wx/ don't > actually include any actual APRS-aware location, but only > an NWS office > designator. You'll note that the t/n (NWS weather > type) filter isn't > supported in the .../call/range form. So, there's no > defined way to get > "local" NWS alerts from the APRS-IS feed for gating to > start with. The > filters are all-or-nothing. Not quite, which is why I was pondering an actual NWS gate independent of regular Igate software. All the messages have the zone data in them which does provide geographic information. The idea was to have the NWSgate understand which zones are local and which are not (not a difficult task) via the configuration. Some of the messages also contain the FIPS codes which are also useful. As an example, if I were still in Atlanta, I would have designated the various counties that make up the Atlanta and surrounding areas (about thirteen counties) but, due to the prevailing winds, I would also designate the counties west of Atlanta leading into the first few counties of Alabama. Having those zone codes in the config file, the gate could then send out to RF only those watches and warnings that lit up those counties with advance notice (from the western counties) plus any affecting the current counties. I don't need the eastern counties outside of the Igate footprint because the storm is passing the Igate by that point. > > I'm trying to figure out a way to support a filtered gating > of such > information for any IGates running APRSISCE/32 and they'll > probably be > based on the NWS office being explicitly specified in the > filter > component(s). _______________________________________________ Xastir mailing list [email protected] http://lists.xastir.org/cgi-bin/mailman/listinfo/xastir
